• Aonashi, K., and G. Liu, 2000: Passive microwave precipitation retrievals using TMI during the baiu period of 1998. Part I: Algorithm description and validation. J. Appl. Meteor., 39, 20242037, doi:10.1175/1520-0450(2000)039<2024:PMPRUT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Aonashi, K., and Coauthors, 2009: GSMaP passive microwave precipitation retrieval algorithm: Algorithm description and validation. J. Meteor. Soc. Japan, 87A, 119136, doi:10.2151/jmsj.87A.119.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., 1996: The effect of vertical shear on tropical cyclone intensity change. J. Atmos. Sci., 53, 20762088, doi:10.1175/1520-0469(1996)053<2076:TEOVSO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., and J. Kaplan, 1994: A Statistical Hurricane Intensity Prediction Scheme (SHIPS) for the Atlantic basin. Wea. Forecasting, 9, 209220, doi:10.1175/1520-0434(1994)009<0209:ASHIPS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., and J. Kaplan, 1999: An updated Statistical Hurricane Intensity Prediction Scheme (SHIPS) for the Atlantic and eastern North Pacific basins. Wea. Forecasting, 14, 326337, doi:10.1175/1520-0434(1999)014<0326:AUSHIP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., M. Mainelli, L. K. Shay, J. A. Knaff, and J. Kaplan, 2005: Further improvements to the Statistical Hurricane Intensity Prediction Scheme (SHIPS). Wea. Forecasting, 20, 531543, doi:10.1175/WAF862.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1986: An air–sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43, 585605, doi:10.1175/1520-0469(1986)043<0585:AASITF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hendricks, E. A., M. S. Peng, B. Fu, and T. Li, 2010: Quantifying environmental control on tropical cyclone intensity change. Mon. Wea. Rev., 138, 32433271, doi:10.1175/2010MWR3185.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holliday, C. R., and A. H. Thompson, 1979: Climatological characteristics of rapidly intensifying typhoons. Mon. Wea. Rev., 107, 10221034, doi:10.1175/1520-0493(1979)107<1022:CCORIT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • JAXA, 2016a: JAXA global rainfall watch. Accessed 18 November 2016. [Available online at http://sharaku.eorc.jaxa.jp/GSMaP/index.htm.]

  • JAXA, 2016b: JAXA realtime rainfall watch. Accessed 18 November 2016. [Available online at http://sharaku.eorc.jaxa.jp/GSMaP_NOW/index.htm.]

  • Jiang, H., 2012: The relationship between tropical cyclone intensity change and the strength of inner-core convection. Mon. Wea. Rev., 140, 11641176, doi:10.1175/MWR-D-11-00134.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jiang, H., and E. M. Ramirez, 2013: Necessary conditions for tropical cyclone rapid intensification as derived from 11 years of TRMM data. J. Climate, 26, 64596470, doi:10.1175/JCLI-D-12-00432.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jones, S. C., 1995: The evolution of vortices in vertical shear. I: Initially barotropic vortices. Quart. J. Roy. Meteor. Soc., 121, 821851, doi:10.1002/qj.49712152406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joyce, R. J., J. E. Janowiak, P. A. Arkin, and P. Xie, 2004: CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution. J. Hydrometeor., 5, 487503, doi:10.1175/1525-7541(2004)005<0487:CAMTPG>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kaplan, J., and M. DeMaria, 2003: Large-scale characteristics of rapidly intensifying tropical cyclones in the North Atlantic basin. Wea. Forecasting, 18, 10931108, doi:10.1175/1520-0434(2003)018<1093:LCORIT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kaplan, J., M. DeMaria, and J. A. Knaff, 2010: A revised tropical cyclone rapid intensification index for the Atlantic and eastern North Pacific basins. Wea. Forecasting, 25, 220241, doi:10.1175/2009WAF2222280.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kieper, M., and H. Jiang, 2012: Predicting tropical cyclone rapid intensification using the 37 GHz ring pattern identified from passive microwave measurements. Geophys. Res. Lett., 39, L13804, doi:10.1029/2012GL052115.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kieu, C., V. Tallapragada, and W. Hogsett, 2014: Vertical structure of tropical cyclones at onset of the rapid intensification in the HWRF model. Geophys. Res. Lett., 41, 32983306, doi:10.1002/2014GL059584.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kubota, T., and Coauthors, 2007: Global precipitation map using satellite-borne microwave radiometers by the GSMaP project: Production and validation. IEEE Trans. Geosci. Remote Sens., 45, 22592275, doi:10.1109/TGRS.2007.895337.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kubota, T., S. Shige, K. Aonashi, and K. Okamoto, 2009: Development of nonuniform beamfilling correction method in rainfall retrievals for passive microwave radiometers over ocean using TRMM observations. J. Meteor. Soc. Japan, 87A, 153164, doi:10.2151/jmsj.87A.153.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miyamoto, Y., and T. Takemi, 2013: A transition mechanism for the axisymmetric spontaneous intensification of tropical cyclones. J. Atmos. Sci., 70, 112129, doi:10.1175/JAS-D-11-0285.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miyamoto, Y., and T. Takemi, 2015: A triggering mechanism for rapid intensification of tropical cyclones. J. Atmos. Sci., 72, 26662681, doi:10.1175/JAS-D-14-0193.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molinari, J., and D. Vollaro, 2010: Rapid intensification of a sheared tropical storm. Mon. Wea. Rev., 138, 38693885, doi:10.1175/2010MWR3378.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molinari, J., D. Vollaro, and K. L. Corbosiero, 2004: Tropical cyclone formation in a sheared environment: A case study. J. Atmos. Sci., 61, 24932509, doi:10.1175/JAS3291.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nolan, D. S., and L. D. Grasso, 2003: Nonhydrostatic, three-dimensional perturbations to balanced, hurricane-like vortices. Part II: Symmetric response and nonlinear simulations. J. Atmos. Sci., 60, 27172745, doi:10.1175/1520-0469(2003)060<2717:NTPTBH>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nolan, D. S., Y. Moon, and D. P. Stern, 2007: Tropical cyclone intensification from asymmetric convection: Energetics and efficiency. J. Atmos. Sci., 64, 33773405, doi:10.1175/JAS3988.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pendergrass, A. G., and H. E. Willoughby, 2009: Diabatically induced secondary flows in tropical cyclones. Part I: Quasi-steady forcing. Mon. Wea. Rev., 137, 805821, doi:10.1175/2008MWR2657.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reasor, P. D., and M. T. Montgomery, 2001: Three-dimensional alignment and corotation of weak, TC-like vortices via linear vortex Rossby waves. J. Atmos. Sci., 58, 23062330, doi:10.1175/1520-0469(2001)058<2306:TDAACO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reasor, P. D., M. T. Montgomery, and L. D. Grasso, 2004: A new look at the problem of tropical cyclones in vertical shear flow: Vortex resiliency. J. Atmos. Sci., 61, 322, doi:10.1175/1520-0469(2004)061<0003:ANLATP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reasor, P. D., R. Rogers, and S. Lorsolo, 2013: Environmental flow impacts on tropical cyclone structure diagnosed from airborne Doppler radar composites. Mon. Wea. Rev., 141, 29492969, doi:10.1175/MWR-D-12-00334.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Riemer, M., M. T. Montgomery, and M. E. Nicholls, 2010: A new paradigm for intensity modification of tropical cyclones: Thermodynamic impact of vertical wind shear on the inflow layer. Atmos. Chem. Phys., 10, 31633188, doi:10.5194/acp-10-3163-2010.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rogers, R. F., P. Reasor, and S. Lorsolo, 2013: Airborne Doppler observations of the inner-core structural differences between intensifying and steady-state tropical cyclones. Mon. Wea. Rev., 141, 29702991, doi:10.1175/MWR-D-12-00357.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rogers, R. F., P. Reasor, and J. Zhang, 2015: Multiscale structure and evolution of Hurricane Earl (2010) during rapid intensification. Mon. Wea. Rev., 143, 536562, doi:10.1175/MWR-D-14-00175.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schubert, W. H., and J. J. Hack, 1982: Inertial stability and tropical cyclone development. J. Atmos. Sci., 39, 16871697, doi:10.1175/1520-0469(1982)039<1687:ISATCD>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shapiro, L., and H. E. Willoughby, 1982: The response of balanced hurricanes to local sources of heat and momentum. J. Atmos. Sci., 39, 378394, doi:10.1175/1520-0469(1982)039<0378:TROBHT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shige, S., and Coauthors, 2009: The GSMaP precipitation retrieval algorithm for microwave sounders. Part I: Over-ocean algorithm. IEEE Trans. Geosci. Remote Sens., 47, 30843097, doi:10.1109/TGRS.2009.2019954.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stern, D. P., J. L. Vigh, D. S. Nolan, and F. Zhang, 2015: Revisiting the relationship between eyewall contraction and intensification. J. Atmos. Sci., 72, 12831306, doi:10.1175/JAS-D-14-0261.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ushio, T., and Coauthors, 2009: A Kalman filter approach to the Global Satellite Mapping of Precipitation (GSMaP) from combined passive microwave and infrared radiometric data. J. Meteor. Soc. Japan, 87A, 137151, doi:10.2151/jmsj.87A.137.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vigh, J. L., and W. H. Schubert, 2009: Rapid development of the tropical cyclone warm core. J. Atmos. Sci., 66, 33353350, doi:10.1175/2009JAS3092.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zagrodnik, J. P., and H. Jiang, 2014: Rainfall, convection, and latent heating distributions in rapidly intensifying tropical cyclones. J. Atmos. Sci., 71, 27892809, doi:10.1175/JAS-D-13-0314.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Time evolution of γ [hourly (black) and 6-h mean (red)], central pressure (blue), and maximum sustained wind (green) of Typhoon Noul (2015). The orange shading indicates the period when the 6-h mean γ was greater than 0.7 for at least 12 consecutive hours. The vertical purple lines indicate the time of the GSMaP images shown in (b) and (c). (b) GSMaP image and the value of γ at 0400 UTC 8 May 2015. The circle is centered at the TC’s center and has a 100-km radius. A gray background indicates data derived from microwave satellite data, and a white background indicates data derived by interpolation. (c) As in (b), but at 0500 UTC 8 May 2015. (d) As in (a), but for Typhoon Dolphin (2015). (e) As in (b), but at 1000 UTC 15 May 2015. (f) As in (b), but at 1600 UTC 15 May 2015.

  • View in gallery

    (a) As in Fig. 1a, but for Hurricane Earl (2010). (b) As in Fig. 1a, but for Hurricane Igor (2010).

  • View in gallery

    Frequency distribution (number of samples, colors and contours) of the current intensity (horizontal axis) and γ (vertical axis) during the development stage: (a) central pressure and (b) maximum sustained wind. The black curve shows the averaged γ at each current intensity. Samples in (a) are as in Fig. 5b.

  • View in gallery

    (a) Scatter diagram of the intensification rate at t = 0 h vs γ during the development stage. (b) As in (a), but for the decay stage. (c) Correlations between the intensification rate at t − 6, t = 0, t + 6, t + 12, and t + 18 h, and γ at each current intensity during the development stage. The dashed line is drawn at zero. (d) As in (c), but at t − 6, t = 0, and t + 6 h during the decay stage. Note that the dataset used here includes temporary weakening periods during the development stage.

  • View in gallery

    Intensification rate (hPa h−1, colors and contours) relative to the current intensity (central pressure, horizontal axis) and γ (vertical axis) during the development stage: (a) t − 6, (b) t = 0, (c) t + 12, and (d) t + 24 h. Grids containing five or more cases are colored.

  • View in gallery

    Intensity changes in the next 24 h (colors and contours) relative to the current intensity (horizontal axis) and γ (vertical axis) during the development stage: (a) central pressure and (b) maximum sustained wind. Grids containing five or more cases are colored.

  • View in gallery

    Correlations between the intensity change in the next 12, 24, 36, and 48 h, and γ at each current intensity during the development stage: (a) central pressure and (b) maximum sustained wind. Note that the dataset used here includes temporary weakening periods during the development stage. The dashed line is drawn at zero.

  • View in gallery

    Intensity changes (central pressure, colors and contours) in the next 24 h relative to (a) the 6-h mean axisymmetric (horizontal axis) and asymmetric (vertical axis) terms of the rainfall rate, (b) the current intensity (horizontal axis) and the 6-h mean axisymmetric term of the rainfall rate (vertical axis), and (c) the current intensity (horizontal axis) and the 6-h mean asymmetric term of rainfall rate (vertical axis), during the development stage. Only samples with a current intensity from 945 to 995 hPa are included in (a). The values of the axisymmetric term have been normalized to 19.1 mm h−1, and those of the asymmetric term have been normalized to 10.6 mm h−1. Grids containing five or more cases are colored.

  • View in gallery

    (a) Frequency distribution (number of samples, colors and contours) of RI cases relative to the current intensity (central pressure, horizontal axis) and the 24-h mean γ (vertical axis). The black line indicates γ averaged over each current intensity. (b) As in (a), but of non-RI cases. (c) Correlations between the 24-h mean γ and the 24-h intensity change with a time lag of 0 (i.e., from t − 12 to t + 12 h), +6, +12, and +18 h at each current intensity during the development stage. Note that the dataset used here includes temporary weakening periods during the development stage. The dashed line is drawn at zero. (d) Scatter diagram of intensity changes from t − 6 to t + 18 h vs the 24-h mean γ. Only samples with a current intensity from 945 to 995 hPa are included. The red line indicates the average value of γ at each intensity change.

  • View in gallery

    The frequency distributions of 24-h central pressure change (ΔP24) stratified by TC intensity at t = 0 h. The distributions are provided for tropical depressions, tropical storms (both tropical storms and severe tropical storms), typhoons, and all TCs.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 270 127 5
PDF Downloads 218 98 8

Tropical Cyclone Intensity Change and Axisymmetricity Deduced from GSMaP

View More View Less
  • 1 Meteorological Research Institute, Tsukuba, Ibaraki, Japan
  • | 2 Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida, and RIKEN Advanced Institute for Computational Science, Kobe, Hyogo, Japan
© Get Permissions
Full access

Abstract

The relationship of tropical cyclone (TC) future intensity change to current intensity and current axisymmetricity deduced from hourly Global Satellite Mapping of Precipitation (GSMaP) data was investigated. Axisymmetricity is a metric that correlates positively with the magnitude of the axisymmetric component of the rainfall rate and negatively with the magnitude of the asymmetric component. The samples used were all of the TCs that existed in the western North Pacific basin during the years 2000–15. The results showed that, during the development stage, the intensification rate at the current time, and 6 and 12 h after the current time was strongly related to both the current intensity and axisymmetricity. On average, the higher the axisymmetricity, the larger the intensity change in the next 24 h for TCs with a current central pressure (maximum sustained wind) between 945 and 995 hPa (85 and 40 kt). The mean value of the axisymmetricity for TCs experiencing rapid intensification (RI) was much higher than that for non-RI TCs for current intensities of 960–990 hPa. The new observational evidence for the intensification process presented here is consistent with the findings of previous theoretical studies emphasizing the role of the axisymmetric component of diabatic heating.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: Udai Shimada, ushimada@mri-jma.go.jp

Abstract

The relationship of tropical cyclone (TC) future intensity change to current intensity and current axisymmetricity deduced from hourly Global Satellite Mapping of Precipitation (GSMaP) data was investigated. Axisymmetricity is a metric that correlates positively with the magnitude of the axisymmetric component of the rainfall rate and negatively with the magnitude of the asymmetric component. The samples used were all of the TCs that existed in the western North Pacific basin during the years 2000–15. The results showed that, during the development stage, the intensification rate at the current time, and 6 and 12 h after the current time was strongly related to both the current intensity and axisymmetricity. On average, the higher the axisymmetricity, the larger the intensity change in the next 24 h for TCs with a current central pressure (maximum sustained wind) between 945 and 995 hPa (85 and 40 kt). The mean value of the axisymmetricity for TCs experiencing rapid intensification (RI) was much higher than that for non-RI TCs for current intensities of 960–990 hPa. The new observational evidence for the intensification process presented here is consistent with the findings of previous theoretical studies emphasizing the role of the axisymmetric component of diabatic heating.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author e-mail: Udai Shimada, ushimada@mri-jma.go.jp

1. Introduction

It has recently been revealed that the intensification rate of tropical cyclones (TCs) is not determined by environmental conditions only but is also controlled by dynamical and thermodynamical processes in the TC’s inner core. Hendricks et al. (2010) showed that there is little difference in environmental conditions between TCs that rapidly intensify and those that intensify at a normal rate. They concluded that the intensification rate is controlled mostly by internal dynamical processes, provided that environmental conditions are relatively favorable. Miyamoto and Takemi (2013) performed an idealized numerical simulation that showed the importance of internal processes in achieving rapid intensification (RI). A TC vortex becomes axisymmetric, and inertial stability substantially increases, before the onset of RI. The large inertial stability increases the residence time of air parcels in the inner core, which enhances convective available potential energy (CAPE) within the radius of maximum wind (RMW). Then an eyewall forms by consuming the enhanced CAPE, and the intensification rate rapidly increases. Kieu et al. (2014) performed idealized numerical simulations and hypothesized that an upper-level warm core, high humidity in the inner core, and sufficient low-level tangential flow were necessary for the onset of RI. The findings of these studies suggest that no matter how favorable environmental conditions are, a TC cannot intensify unless favorable internal conditions have been realized.

Theoretical studies have examined the mechanisms of TC vortex intensification in terms of the response to heat sources. In the case of a radially located heat source, Pendergrass and Willoughby (2009) and Shapiro and Willoughby (1982) have demonstrated that diabatic heating inside the RMW leads to a rapid increase in tangential wind speed. They also showed that eyewall heating causes more rapid strengthening of the tangential wind as the maximum wind speed increases. Vigh and Schubert (2009) and Schubert and Hack (1982) found that diabatic heating inside the RMW is more efficient for the formation of a warm core than diabatic heating outside the RMW. When the heat source is distributed azimuthally, Nolan and Grasso (2003) and Nolan et al. (2007) showed that the axisymmetric component of diabatic heating is much more efficient for the vortex intensification than the asymmetric component; in fact, asymmetric heating can contribute directly to the weakening of a TC vortex. Nolan et al. (2007) also showed that the ratio of energy retained as wind kinetic energy to the injected heat energy increases as the maximum wind speed increases.

In an observational study, Rogers et al. (2013) performed composite analyses of airborne Doppler measurements to compare inner-core structural differences between intensifying and steady-state hurricanes. They showed that intensifying hurricanes have a ringlike vorticity structure inside the RMW and greater azimuthal precipitation coverage in the eyewall compared with steady-state hurricanes. Zagrodnik and Jiang (2014) investigated the distribution of rainfall and diabatic heating by using Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar (PR) data and showed that both distributions become more axisymmetric immediately before the onset of RI, whereas the rainfall area extends farther on the upshear side during RI. Kieper and Jiang (2012) reported that the existence of an eyewall ring on the Naval Research Laboratory (NRL) 37-GHz passive microwave composite product that consists of low-level water clouds and warm rain is important for RI. Meanwhile, Jiang (2012) used TRMM data to show that strong convection is neither a necessary nor a sufficient condition for RI, although they found a statistically significant relationship between convective intensity in the inner core and TC intensity changes. Also using TRMM data, Jiang and Ramirez (2013) found that rainfall in the inner core is at least moderate to heavy (i.e., total raining area >3000 km2, total volumetric water >5000 mm h−1 km2) and convection in the inner core is at least moderate [i.e., maximum near-surface radar reflectivity >40 dBZ, maximum 20-dBZ (40 dBZ) echo height >8 (4) km, minimum 85-GHz polarization-corrected brightness temperature (PCT) <235 K, and minimum 10.8-mm brightness temperature <220 K] in TCs before the onset of RI, but intense asymmetric convection is not a necessary condition for the onset of RI. In contrast, Rogers et al. (2013) showed in composite analyses that intensifying hurricanes are characterized by convective bursts (deep, vigorous convection) that preferentially occur inside the RMW. Rogers et al. (2015), who used airborne observations to investigate Hurricane Earl (2010), also emphasized the role of asymmetric convective bursts during RI.

The results of these observational studies contain two views about conditions favorable for intensification from the perspective of azimuthal distributions: the axisymmetric component of diabatic heating is important, as is asymmetric intense convection. Although relationships between inner-core structures and TC intensity changes have been investigated both theoretically and observationally, the relative contribution of axisymmetric and asymmetric components to intensity changes still remains a matter of debate. Further clarification is needed.

In this study, we investigated the relationship between TC intensity changes and axisymmetricity deduced from hourly Global Satellite Mapping of Precipitation (GSMaP) data (Kubota et al. 2007, 2009), a satellite-derived rainfall product. We defined axisymmetricity as a metric positively correlated with the magnitude of the axisymmetric component of the rainfall rate and negatively correlated with the magnitude of the asymmetric component (see section 2). Because the magnitude of the rainfall rate can be assumed to be correlated with the magnitude of diabatic heating in the low- to midlevel troposphere, this metric is hypothesized to have a strong relationship with the TC intensity change. Previous studies using satellite data have investigated TC features in relation to RI, whereas in this study we examined features in relation to differences in the intensification rate.

This paper consists of five sections. We describe the data used and the method in section 2. In section 3, we present the results of the investigation. In section 4, we discuss the results with a focus on the relative relationships between axisymmetric and asymmetric components and intensity changes. Section 5 is a summary.

2. Data and method

We used hourly GSMaP data composed of high-resolution (0.1°) satellite-based rainfall rate (mm h−1) estimates. These data are processed and provided by the Japan Aerospace Exploration Agency (JAXA) (JAXA 2016a). The spatially and temporally high-resolution GSMaP dataset is constructed in two steps.

In the first step, the rainfall rate is estimated by a retrieval algorithm developed by Aonashi and Liu (2000), Aonashi et al. (2009), and Shige et al. (2009) from microwave satellite data from the TRMM Microwave Imager (TMI), the Advanced Earth Observing Satellite-II (ADEOS-II) Advanced Microwave Scanning Radiometer (AMSR), the Aqua AMSR for Earth Observing System (AMSR-E), the Global Change Observation Mission–Water 1 (GCOM-W1) AMSR2, the Defense Meteorological Satellite Program (DMSP) Special Sensor Microwave Imager/Sounder (SSMIS) series, the National Oceanic and Atmospheric Administration (NOAA) Advanced Microwave Sounding Unit series (AMSU), and the Meteorological Operational Satellite Program of Europe (MetOp) AMSU series, the Global Precipitation Measurement Core Observatory (GPM-Core) Microwave Imager (GMI). The use of these microwave satellite data enables the satellite-retrieved rainfall dataset to have a temporal resolution of about 3–6 h at each point and to cover the area from 60°S to 60°N; however, the rainfall rate is not retrieved in areas of expected surface temperatures below 0°C, as snowfall is expected.

In the second step, hourly rainfall rate estimates are obtained from the satellite-retrieved rainfall data by means of two temporal interpolation methods. The first is the morph method (Joyce et al. 2004), in which the movement of precipitation areas retrieved from microwave satellite data is estimated by using migration vectors derived from geostationary satellite infrared (IR) imagery. The second is rainfall correction by a Kalman filter, which uses the statistical relationship between geostationary satellite IR brightness temperatures (TBs) and precipitation (Ushio et al. 2009). Details of the GSMaP dataset are given by Kubota et al. (2009) and the GSMaP website (JAXA 2016a). Although the microwave satellites used to obtain the GSMaP data orbit Earth so as to generally cover the entire GSMaP region at intervals of no more than 3–6 h, quality degradation due to the temporal interpolation of precipitation may affect the axisymmetricity of TCs. We will quantify this impact in section 3a.

GSMaP data are available since 2000. In this study, we used the GSMaP reanalysis product (version 6 algorithm without the gauge correction) for the period from 2000 to 2015.

We used the metric of axisymmetricity γ, defined by Miyamoto and Takemi (2013). First, a variable is divided into the azimuthal mean (axisymmetric component, denoted by an overbar) and the deviation from the azimuthal mean (asymmetric component, denoted by a prime):
e1
where r, , and z are the radial, tangential, and axial direction, and t is time, respectively. Axisymmetricity is then defined as follows:
e2
Here γ depends on the magnitude of the axisymmetric component as well as on that of the asymmetric component. The value γ lies between 0 and 1, and is largest when the asymmetric component is 0 all the way around the circle.

Although Miyamoto and Takemi (2013) used potential vorticity (PV) as in their definition of axisymmetricity, we used the rainfall rate from the GSMaP dataset. First γ is calculated at each radius, and then γ is averaged from the TC center to the 100-km radius, in the same way as in Miyamoto and Takemi (2013). We chose 100 km as the outermost radius from which to obtain the mean axisymmetricity in the inner core of a TC because the RMW is smaller than 100 km for most TCs (e.g., Stern et al. 2015). Miyamoto and Takemi (2013) showed that a distinct eyewall forms and RI begins several hours after the distribution of PV becomes axisymmetric. The use of precipitation data instead of PV might make it difficult to capture the axisymmetrization of TCs several hours before the start of intensification in a similar way to the PV field.

We prepared a dataset for this study consisting of all 380 TCs that occurred in the western North Pacific (WNP) basin (from 0° to 60°N and from 100°E to 180°) from 2000 to 2015. Because the temporal resolution of the GSMaP data is hourly, we could obtain a sufficient number of samples for the investigation in the WNP basin. To obtain TC centers and current intensities [in terms of both central pressure and maximum sustained wind (10-min mean)], we used the best track data of the Regional Specialized Meteorological Center (RSMC) Tokyo, Japan, which were linearly interpolated to 1-h intervals. Although the GSMaP data can be used to investigate TCs all over the world, which is one of the advantages of this dataset, we mainly analyzed TCs in the WNP basin because the success of our approach depends greatly on the quality of the best track data used. The TC dataset included TCs classified as tropical storms (maximum sustained wind speed, 34–47 kt, 1 kt = 0.5144 m s−1), severe tropical storms (48–63 kt), and typhoons (>64 kt) in the best track data, and excluded TCs during the tropical depression stage (<34 kt), because the best track data did not provide maximum sustained wind data during that stage. However, if a TC decayed into a tropical depression and then reintensified into a tropical storm, this tropical depression period was, exceptionally, included in the dataset to simplify data handling.

As intensity, we evaluated both central pressure and maximum sustained wind speed. We used knots (kt) as the units for maximum sustained wind speed because the best track data are reported in knots. We investigated the intensification rate (a short-term intensity change) at different times as well as the intensity change in the next 12, 24, 36, and 48 h (a long-term intensity change) to determine causality between γ and TC intensification. The intensification rate was calculated as the centered difference between 3 h before and 3 h after each time point, because most of intensities in the best track data used were available at 6-h intervals. We performed the investigation by dividing the TC data according to two TC stages: the development stage and the decay stage. The development stage was defined as a period from the time that the TC was first classified as a tropical storm to the latest time that the TC reached its maximum intensity during its life cycle. However, data in which the change of central pressure (maximum sustained wind) was positive (negative) owing to temporary weakening and landfall were excluded from the development stage dataset. The decay stage was defined as a period from 1 h after the development stage to the last time that the TC was classified as a tropical storm, including periods of reintensification and landfall. In the case of TCs that entered the WNP basin from the central Pacific basin, we determined both the development and decay stages for the entire period covered by the best track data (i.e., while the TC was in the WNP basin).

Finally, we binned the dataset into 5-hPa and 5-kt intervals of current intensity and 0.02 intervals of current γ.

3. Results

a. Case study

First, we examined whether γ was related to intensity changes in specific cases, while quantifying uncertainties in γ deduced from the hourly GSMaP data. To exclude the possibility that the results obtained were strongly dependent on the best track data, we selected examples in the North Atlantic basin, along with best track data provided by the National Hurricane Center (NHC), as well as examples from the WNP basin. The NHC’s best track data, which are based on in situ observations by aircraft and obtained by dropsonde as well as on satellite data, are expected to be more reliable than data from other centers. Note that the maximum sustained wind speed in the NHC’s best track is a 1-min average, whereas that in the RSMC Tokyo’s best track is a 10-min average.

We examined two typhoons, Typhoon Noul (2015) and Typhoon Dolphin (2015) (Fig. 1), and two hurricanes, Hurricane Earl (2010) and Hurricane Igor (2010) (Fig. 2). In the time evolution of each case, hourly γ showed erratic fluctuations. These fluctuations had several causes: 1) inaccurate TC center locations, when the TC center interpolated from the best track data deviated from the actual location (Fig. 1e); 2) different data resolutions among microwave satellites (Figs. 1e,f); 3) quality differences due to the use of the temporal interpolation method (Figs. 1b,c); and 4) biases in rainfall-rate estimates among microwave satellites. We quantified uncertainties in γ associated with these causes (Table 1 and 2). As for the first cause of the erratic fluctuations, we evaluated the uncertainties from the perspective of the degree to which differences in center locations can vary γ. For each center in Figs. 1b, 1c, 1e, and 1f, we prepared a total of 24 centers around the original center regularly spaced 0.1° apart and filled a 0.4° × 0.4° square and then we calculated γ one by one to estimate the standard deviation of 25 γ values (including original γ). The results showed that the standard deviation ranged from 0.02 to 0.11 (Table 1). For the second cause, there were systematic differences in γ reaching 0.25 between imager and sounder sensors (Table 2). For the third cause, γ tended to be higher when the temporal interpolation by IR data was applied (Table 2). For the fourth cause, there were differences in γ even among imager sensors and among sounder sensors (Table 2). Fortunately, when γ was averaged over 6 h (an average between 3 h before and 3 h after each time point), these finescale fluctuations disappeared and there remained a time evolution of γ that fluctuated with a period of a few hours (Figs. 1a,d). When the hourly γ was replaced by the 6-h mean γ, differences in γ among sensors almost disappeared (Table 2). Although the problems of data quality and TC center deviations are limitations of this study that should be addressed in the future, we decided to use 6-h mean values of γ to examine the relationships between future TC intensity changes and current values of intensity and γ.

Fig. 1.
Fig. 1.

(a) Time evolution of γ [hourly (black) and 6-h mean (red)], central pressure (blue), and maximum sustained wind (green) of Typhoon Noul (2015). The orange shading indicates the period when the 6-h mean γ was greater than 0.7 for at least 12 consecutive hours. The vertical purple lines indicate the time of the GSMaP images shown in (b) and (c). (b) GSMaP image and the value of γ at 0400 UTC 8 May 2015. The circle is centered at the TC’s center and has a 100-km radius. A gray background indicates data derived from microwave satellite data, and a white background indicates data derived by interpolation. (c) As in (b), but at 0500 UTC 8 May 2015. (d) As in (a), but for Typhoon Dolphin (2015). (e) As in (b), but at 1000 UTC 15 May 2015. (f) As in (b), but at 1600 UTC 15 May 2015.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

Fig. 2.
Fig. 2.

(a) As in Fig. 1a, but for Hurricane Earl (2010). (b) As in Fig. 1a, but for Hurricane Igor (2010).

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

Table 1.

Uncertainty evaluations of γ due to inaccurate TC center locations. Values of γ calculated by using the original center, and mean and standard deviation (std dev) γ calculated by using the 25 centers (the 24 centers plus the original center) are provided. See text for details.

Table 1.
Table 2.

Differences in γ among satellites used in the GSMaP data. Here the hourly γ is classified according to which microwave satellite data mainly (>50%) covered a 2° × 2° square area around a TC. Cases in which multiple microwave satellite covered the area simultaneously are excluded. The number of samples (N), mean and standard deviation (std dev) hourly γ, and mean and standard deviation of the 6-h mean γ at each satellite sensor are provided. The mean and standard deviation of the 6-h mean γ at each satellite sensor are computed by replacing the hourly γ with the 6-h mean γ. MHS stands for Microwave Humidity Sounder.

Table 2.

For the four cases shown in Figs. 1 and 2, during the development stage, the maximum sustained wind speed (central pressure) largely increased (decreased) when the 6-h mean γ (hereafter referred to as simply γ) persisted above 0.7 for more than 12 h, whereas the maximum sustained wind speed (central pressure) stopped increasing (decreasing) when γ temporarily decreased to less than 0.7. Although Rogers et al. (2015) reported that asymmetries, including convective bursts, persisted during RI of Hurricane Earl, γ had in fact exceeded 0.7 immediately before the onset of RI (Fig. 2a). In the cases of Hurricane Earl and Igor, minor intensity changes such as those due to eyewall replacement cycles after RI showed little correlation with changes of γ. During the decay stage, γ was not related to intensity changes.

Because the relationship between intensity changes and γ during the development stage was seen in TCs not only in the WNP basin but also in the North Atlantic basin, depending on the current intensity, the relationship was not likely to be dependent on either the basin in which TCs occur or the quality of the best track data used.

b. Relationship between TC intensity changes and axisymmetricity

In this subsection, we first show the relationship between current intensity and γ, and the relationship between future intensity change and γ. Then, we present the relative contribution of axisymmetric and asymmetric components of the rainfall rate to intensification. Finally, we present the relationship between RI and γ.

1) Current intensity versus axisymmetricity

The frequency distribution of current intensity and γ during the development stage showed a peak around current intensities of 990 hPa or 50 kt and γ of ~0.73 (Fig. 3). Roughly speaking, averaged γ at each current intensity showed that within the current intensity range of 1000–955 hPa (35–90 kt), γ increased with the current intensity, which implies that within these intensity ranges, the structure of many TCs changed from asymmetric to axisymmetric. In addition to this structural change, an increase in the azimuthally averaged rainfall rate contributed to the increase of γ at around these intensities. In contrast, at central pressures <955 hPa (maximum sustained wind >90 kt), mean γ was almost constant. The number of cases with current intensity <920 hPa (>110 kt) and the number of cases with γ greater than 0.95 were small. In addition, γ ranged widely at each current intensity.

Fig. 3.
Fig. 3.

Frequency distribution (number of samples, colors and contours) of the current intensity (horizontal axis) and γ (vertical axis) during the development stage: (a) central pressure and (b) maximum sustained wind. The black curve shows the averaged γ at each current intensity. Samples in (a) are as in Fig. 5b.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

2) Future intensity change versus axisymmetricity

Next, we examined the intensification rate at different times in relation to the current intensity and γ during the development stage. Here, we included temporary weakening periods during the development stage in the dataset of Figs. 4a and 4c to calculate correlations between the intensification rate and γ. The scatter diagram of the intensification rate at the current time (t = 0 h) and γ showed that the intensification rate was strongly related to γ, although the data are widely scatted (Fig. 4a). The intensification rate at t = 0, t + 6, and t + 12 h had relatively high correlations with γ for intensities of 960–1000 hPa (Fig. 4c). In addition, this relationship continued to be strong after t + 12 h for intensities >970 hPa.

Fig. 4.
Fig. 4.

(a) Scatter diagram of the intensification rate at t = 0 h vs γ during the development stage. (b) As in (a), but for the decay stage. (c) Correlations between the intensification rate at t − 6, t = 0, t + 6, t + 12, and t + 18 h, and γ at each current intensity during the development stage. The dashed line is drawn at zero. (d) As in (c), but at t − 6, t = 0, and t + 6 h during the decay stage. Note that the dataset used here includes temporary weakening periods during the development stage.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

The scatter diagram during the decay stage showed that, compared with the development stage, the data were displaced in the x direction by about +1 hPa h−1 with a wider spread (Fig. 4b). Reintensifying TCs had relatively high γ (Fig. 4b) and the decay rate was slightly correlated with γ if including reintensifying TCs (Fig. 4d). However, within only weakening TCs there was little relationship between the decay rate and γ (Fig. 4b).

Figure 5 shows the relationship between the intensification rate at different times and current intensity and γ. On average, as γ became higher, the intensification rate at t = 0 and t + 12 h tended to increase for TCs with current intensities of 945–995 hPa (Figs. 5b and 5c) (85–40 kt, not shown). This relationship, though less pronounced, could also be seen at t − 6 h, but at intensities <945 hPa (>85 kt, not shown) γ was uniformly relatively high (Fig. 5a). These findings show that when TCs finally reached their maximum intensity γ was relatively high because intensity and γ increased together. In other words, the intensification rate of most TCs that reached strong intensities <945 hPa was relatively large at t − 6 h regardless of γ. Consistent with the correlations in Fig. 4c, the relationship still remained at t + 24 h for current intensities of 970–995 hPa (Fig. 5d) (60–40 kt, not shown).

Fig. 5.
Fig. 5.

Intensification rate (hPa h−1, colors and contours) relative to the current intensity (central pressure, horizontal axis) and γ (vertical axis) during the development stage: (a) t − 6, (b) t = 0, (c) t + 12, and (d) t + 24 h. Grids containing five or more cases are colored.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

Because during the development stage the intensification rate at t = 0, t + 12, and t + 24 h was strongly related to both current intensity and γ, there was also a strong relationship between the intensity change in the next 24 h and current intensity and γ (Fig. 6). This relationship was particularly clear for TCs with current central pressures (maximum winds) between 945 and 995 hPa (85 and 40 kt) during the development stage. These intensity ranges are, in general, much weaker than the maximum potential intensity (MPI; e.g., Emanuel 1986; DeMaria and Kaplan 1994). Correlations were also relatively high for the intensity change in the next 12 and 24 h (Fig. 7). For intensities >965 hPa and <70 kt, the relationship of the intensity change in the next 36 and 48 h with current intensity and γ continued to be strong, as long as a TC was still in the development stage (Fig. 7), because the intensification rate at t + 36 h was still strongly related to the current intensity and γ (not shown). These results suggest that once a TC becomes axisymmetric, it can keep its axisymmetric structure and can continue to intensify provided that a favorable environment for intensification exists.

Fig. 6.
Fig. 6.

Intensity changes in the next 24 h (colors and contours) relative to the current intensity (horizontal axis) and γ (vertical axis) during the development stage: (a) central pressure and (b) maximum sustained wind. Grids containing five or more cases are colored.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

Fig. 7.
Fig. 7.

Correlations between the intensity change in the next 12, 24, 36, and 48 h, and γ at each current intensity during the development stage: (a) central pressure and (b) maximum sustained wind. Note that the dataset used here includes temporary weakening periods during the development stage. The dashed line is drawn at zero.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

3) Relative contribution of axisymmetric and asymmetric components to intensification

Because γ used in this study depends not only on the magnitude of the asymmetric component () of the rainfall rate but also on that of the axisymmetric component () [Eq. (2)], we examined how these components contributed to intensification. Here, the asymmetric term, as a representative of the asymmetric component, was calculated as follows:
e3
We normalized values of the axisymmetric term (same as the axisymmetric component) to 19.1 mm h−1, and those of the asymmetric term to 10.6 mm h−1, determined as 98th percentiles of the distributions of axisymmetric and asymmetric values during the development stage to better illustrate grids containing 5 or more cases in Fig. 8. The distribution of intensity changes in the next 24 h relative to the 6-h mean axisymmetric and asymmetric terms in cases with a current intensity from 945 to 995 hPa shows that the magnitudes of the asymmetric and axisymmetric terms generally increased concurrently (Fig. 8a). When the normalized axisymmetric term ranges from 0.1 to 0.7 and the normalized asymmetric term ranges from 0.1 to 0.6, in general, the larger the axisymmetric term and the smaller the asymmetric term were, the larger the intensity change was. The intensity changes in the next 24 h relative to current intensity (central pressure) and the axisymmetric term of the rainfall rate during the development stage (Fig. 8b) show that the axisymmetric term of the rainfall rate and the intensity change were strongly related for current intensities of 935–980 hPa. It is also notable here that at a given current intensity, the intensity change also became larger as the asymmetric term became larger (Fig. 8c). This result follows from the fact that the asymmetric term increased with an increase in the axisymmetric term, although an increase in the asymmetric term was generally smaller than that in the axisymmetric term (Fig. 8a). Therefore, γ is a useful metric that can express the relative contribution of the axisymmetric and asymmetric terms to intensity changes.
Fig. 8.
Fig. 8.

Intensity changes (central pressure, colors and contours) in the next 24 h relative to (a) the 6-h mean axisymmetric (horizontal axis) and asymmetric (vertical axis) terms of the rainfall rate, (b) the current intensity (horizontal axis) and the 6-h mean axisymmetric term of the rainfall rate (vertical axis), and (c) the current intensity (horizontal axis) and the 6-h mean asymmetric term of rainfall rate (vertical axis), during the development stage. Only samples with a current intensity from 945 to 995 hPa are included in (a). The values of the axisymmetric term have been normalized to 19.1 mm h−1, and those of the asymmetric term have been normalized to 10.6 mm h−1. Grids containing five or more cases are colored.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

Although Fig. 8 does not show if the relative contribution is associated with RI, the findings are similar to the results of Kaplan et al. (2010), who constructed an RI index by investigating differences in two predictors between RI and non-RI cases; one predictor was the percentage of the area between radii of 50 and 200 km from the TC center covered by IR TBs colder than −30°C (PX30), and the other was the standard deviation of IR TBs in the same area (SDBT). They showed that the RI cases had higher PX30 values and lower SDBT values.

4) RI versus axisymmetricity

The strong relationships during the development stage raise the following question: To what extent is γ associated with RI? We therefore investigated the relationship between RI and γ. We defined RI as an intensification rate of at least 30 hPa over a 24-h period. This threshold is the 95th percentile of 24-h central pressure changes for WNP TCs, in the same way that Kaplan and DeMaria (2003) defined RI for Atlantic basin TCs using the maximum wind speed (see the appendix for details). In real cases, even if current environmental conditions and current inner-core structures are very favorable for RI, the favorable environmental conditions may not be maintained for very long. If the vertical wind shear suddenly become strong while the TC is developing, then both γ and the intensification rate will decrease. According to Miyamoto and Takemi (2013, 2015), the time after axisymmetrization of the PV field must be sufficient for the onset of RI. Thus, in real cases, it may be difficult to realize RI if high γ is only transient.

On the basis of this reasoning, we examined the relationship between RI and γ using the 24-h mean γ (i.e., an average between t − 12 and t + 12 h), instead of the 6-h mean γ, though in hindsight. Figure 9c demonstrates that the 24-h mean γ has negative correlations <−0.4 with the 24-h intensity change with a time lag of 0–12 h for current intensities of 960–990 hPa. Then, we compared the frequency distributions of current intensity and the current 24-h mean γ between RI cases and non-RI cases (Figs. 9a and 9b). The 24-h intensity change used to classify each case was calculated as the difference between t − 6 and t + 18 h (i.e., a 6-h lag). Most RI cases (~89%) were distributed between current intensities from 925 to 985 hPa and 24-h mean γ values from 0.7 to 0.9. This result suggests that RI can occur when the current intensity is relatively far from the MPI but it rarely occurs when the current intensity is especially weak, as shown previously by Kaplan and DeMaria (2003). Although many non-RI cases were distributed within the same ranges as the RI cases, the average 24-h mean γ at each current intensity was much higher in RI cases than in non-RI cases for current intensities of 960–990 hPa. A scatter diagram (Fig. 9d) of the 24-h mean γ and intensity changes from t − 6 to t + 18 h in cases with a current intensity from 945 to 995 hPa (the range in which the relationship between intensity changes and γ was large; see Fig. 6) shows that, in general, as γ increased, the intensity change became larger, although the data are widely scattered.

Fig. 9.
Fig. 9.

(a) Frequency distribution (number of samples, colors and contours) of RI cases relative to the current intensity (central pressure, horizontal axis) and the 24-h mean γ (vertical axis). The black line indicates γ averaged over each current intensity. (b) As in (a), but of non-RI cases. (c) Correlations between the 24-h mean γ and the 24-h intensity change with a time lag of 0 (i.e., from t − 12 to t + 12 h), +6, +12, and +18 h at each current intensity during the development stage. Note that the dataset used here includes temporary weakening periods during the development stage. The dashed line is drawn at zero. (d) Scatter diagram of intensity changes from t − 6 to t + 18 h vs the 24-h mean γ. Only samples with a current intensity from 945 to 995 hPa are included. The red line indicates the average value of γ at each intensity change.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

One reason for the wide scatter may be differences in environmental conditions among cases. For example, even if vertical wind shear is weak and the internal conditions of a TC are favorable (i.e., relatively high γ), a TC that has almost reached its MPI cannot intensify further. In addition, TCs that intensify rapidly under relatively strong vertical wind shear with asymmetric structures (i.e., relatively low γ) (e.g., Molinari et al. 2004; Molinari and Vollaro 2010) might contribute to the wide scatter. Moreover, because γ was averaged within a 100-km radius, its value would depend on the horizontal scale of each TC; thus, differences in horizontal scale would also contribute to the wide scatter. Finally, a variety of observational errors by satellites, sensor biases among satellites, analysis errors in the GSMaP, and the best track data could also cause wide scatter.

4. Discussion

The relative relationship between the axisymmetric and asymmetric terms of the rainfall distribution and intensity changes during the development stage was such that the intensity change was larger when the axisymmetric term was larger and the asymmetric term was smaller (Fig. 8a). The axisymmetricity γ of the rainfall distribution, which is a metric that reflects these relative relationships, could be strongly related to subsequent intensity changes, but this relationship was strongly dependent on the current intensity. The results of this study strongly support from an observational perspective the finding of Miyamoto and Takemi (2013) that the axisymmetrization of a TC structure can lead subsequently to significant intensification. In addition, considering that the distribution of the rainfall rate is quite similar to that of diabatic heating (Zagrodnik and Jiang 2014), the results of this study are also very well consistent with the findings of Nolan and Grasso (2003) and Nolan et al. (2007) that the axisymmetric component of diabatic heating is much more important for vortex intensification than the asymmetric component, and that the efficiency of the intensification increases with the vortex strength, provided that the current intensity is well below the MPI. Because the asymmetric component increased along with the axisymmetric component, intense asymmetric convection may also contribute to intensification through an increase in the axisymmetric component of diabatic heating.

Axisymmetricity is an indicator of a TC’s inner-core structure, not an environmental parameter. However, axisymmetry can be changed by environmental vertical shear or by dry air intrusion (e.g., Riemer et al. 2010; Reasor et al. 2013) as well as by internal dynamics. In addition, a TC vortex can have resiliency against vertical shear that helps it retain its axisymmetry (e.g., Reasor and Montgomery 2001; Reasor et al. 2004). The degree of resiliency depends on the inertial stability, that is, TC intensity (e.g., DeMaria 1996; Jones 1995). Although it is difficult to determine if a change of γ should be attributed to internal TC dynamics or to environmental forcing, it might be interesting in a future study to further examine the relationship between γ and the magnitude of vertical wind shear.

Although individual intensification rates were widely scattered in relation to the parameter γ alone, it might be possible to explain intensity changes more adequately by using γ in combination with other parameters in a multiple regression analysis. For example, by adding γ-related predictors to the Statistical Hurricane Intensity Prediction Scheme (SHIPS) model (DeMaria and Kaplan 1994, 1999; DeMaria et al. 2005) and to the RI index (Kaplan et al. 2010) might improve their forecast skill. In our next work we plan to investigate ways to incorporate γ into the SHIPS model and RI index.

For the findings of this study to have operational use, several issues remain to be addressed. The first is the problem of TC center deviation, which caused erratic fluctuations in γ, which we overcame by using the 6-h mean γ. Another problem is the use of GSMaP reanalysis product data, which are not provided in real time. Recently, a real-time GSMaP data product has become available with little latency (JAXA 2016b). Use of this product would resolve the problem of latency for operational use, but at the expense of data quality. The quality might be improved for operational application by using IR TB data from geostationary satellites in combination with the real-time GSMaP data. A future study should examine the feasibility of this approach. The SHIPS model has already incorporated IR TB-based parameters that represent axisymmetricity to some extent [PX30 and SDBT; see section 3b; DeMaria et al. (2005)]. However, it should be noted that IR TBs, which are temperatures at cloud-top heights, cannot be assumed to be correlated with the amount of diabatic heating, whereas the rainfall rate used in this study can.

The results of this study suggest that GSMaP data are useful for elucidating intensification processes of TCs and for diagnosing TC intensity changes for a short-range forecast. The axisymmetricity parameter γ could capture changes of inner-core structure from asymmetry to axisymmetry, but it could not capture convective bursts and structural changes such as eyewall replacement cycles. Our ongoing study will attempt to identify other possible parameters related to intensity changes by combining GSMaP data with microwave and IR satellite data, for example, parameters associated with eyewall replacement cycles and the relationship between the intensification rate and the radial location of convective bursts inside or outside the RMW (Rogers et al. 2013). In this regard, the determination of the RMW itself is a challenging problem.

5. Summary

The relative relationship between axisymmetric and asymmetric components of TC rainfall distribution and TC intensity changes, including during the decay stage, has not been comprehensively investigated in real cases. Based on a study by Miyamoto and Takemi (2013), we investigated the relationship between TC future intensity change during the development and decay stages and both current intensity (both central pressure and maximum sustained wind) and axisymmetricity deduced from hourly GSMaP data. The GSMaP data are satellite-derived rainfall estimates in the region from 60°S to 60°N with a resolution of 0.1°. As defined in this study, axisymmetricity is a metric positively correlated with the magnitude of the axisymmetric component of the rainfall rate and negatively correlated with the magnitude of the asymmetric component. For samples, we used all 380 TCs that existed in the WNP basin from 2000 to 2015. Case studies showed that a TC tended to intensify rapidly when the axisymmetricity persisted at relatively high values for more than 12 h. In general, during the development stage, the intensification rate at t = 0, t + 6, and t + 12 h was strongly related to both the current intensity and axisymmetricity. On average, the higher the axisymmetricity of the TC, the larger the intensity change was in the next 24 h in the case of TCs with current central pressures (maximum sustained winds) between 945 and 995 hPa (85 and 40 kt). In addition, for current intensities >965 hPa and <70 kt, the relationship of the intensity change in the next 36 and 48 h with current intensity and axisymmetricity continued to be strong. These results suggest that once a TC becomes axisymmetric, it can keep its axisymmetric structure and can continue to intensify provided that a favorable environment for intensification exists. During the decay stage, although reintensifying TCs had relatively high axisymmetricity, there was no relationship between the decay rate and axisymmetricity. The mean axisymmetricity of TCs experiencing RI was much higher than that of non-RI TCs for current intensities of 960–990 hPa. The relative relationship between axisymmetric and asymmetric terms showed that the larger the axisymmetric term and the smaller the asymmetric term, the larger the intensity change was, although the asymmetric term generally increased with the increase of the axisymmetric term. The metric of axisymmetricity used in this study could properly express this relative relationship. The new observational evidence presented here is consistent with the findings of previous theoretical studies of the intensification process, which emphasize the role of the axisymmetric component of diabatic heating.

Acknowledgments

U. Shimada is deeply grateful to Dr. M. Yamaguchi, Dr. K. Okamoto, Dr. R. Coronel, Dr. M. Sawada, Ms. H. Owada, and Mr. N. Koide for helpful advice. The authors thank three anonymous reviewers for valuable comments that have greatly improved the manuscript. This work was partly supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) KAKENHI Grant 15K05294, and the Eighth Precipitation Measuring Mission (PMM) of JAXA.

APPENDIX

Definition of RI for WNP TCs

Kaplan and DeMaria (2003) defined RI as the 95th percentile of intensity changes over a 24-h period for Atlantic basin TCs, which is a maximum sustained wind speed increase of 30 kt. For WNP TCs, Holliday and Thompson (1979) used a 42-hPa threshold over a 24-h period. However, in the recent RSMC Tokyo’s best track data (e.g., 2000–15), this threshold can capture very few samples (above the 98.5th percentile).

In this study, we determined an RI threshold using the central pressure from the RSMC Tokyo’s best track data (2000–15, at 6-h intervals) in the same way as in Kaplan and DeMaria (2003). Namely, intensity change was evaluated from t = 0 to t + 24 h after provided that the storm remained overwater and remained a tropical cyclone from t − 12 to t + 24 h. Note that the RSMC Tokyo’s best track data do not include disturbances whose lives ended without evolving into, at least, tropical storms.

Figure A1 shows the cumulative frequency distributions of 24-h central pressure change (ΔP24). The 95th percentile of the ΔP24 distribution was 30 hPa. Details of the samples are shown in Tables A1 and A2.

Fig. A1.
Fig. A1.

The frequency distributions of 24-h central pressure change (ΔP24) stratified by TC intensity at t = 0 h. The distributions are provided for tropical depressions, tropical storms (both tropical storms and severe tropical storms), typhoons, and all TCs.

Citation: Monthly Weather Review 145, 3; 10.1175/MWR-D-16-0244.1

Table A1.

Statistics of central pressure changes over a 24-h period (ΔP24) for tropical depressions, tropical storms, typhoons, and all TCs. The number of samples (N), mean and standard deviation (std dev), minimum (min), and maximum (max) ΔP24 are provided.

Table A1.
Table A2.

The distribution of central pressure change over a 24-h period (ΔP24) for the 437 RI cases. The number of RI cases is stratified by intensity class.

Table A2.

REFERENCES

  • Aonashi, K., and G. Liu, 2000: Passive microwave precipitation retrievals using TMI during the baiu period of 1998. Part I: Algorithm description and validation. J. Appl. Meteor., 39, 20242037, doi:10.1175/1520-0450(2000)039<2024:PMPRUT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Aonashi, K., and Coauthors, 2009: GSMaP passive microwave precipitation retrieval algorithm: Algorithm description and validation. J. Meteor. Soc. Japan, 87A, 119136, doi:10.2151/jmsj.87A.119.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., 1996: The effect of vertical shear on tropical cyclone intensity change. J. Atmos. Sci., 53, 20762088, doi:10.1175/1520-0469(1996)053<2076:TEOVSO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., and J. Kaplan, 1994: A Statistical Hurricane Intensity Prediction Scheme (SHIPS) for the Atlantic basin. Wea. Forecasting, 9, 209220, doi:10.1175/1520-0434(1994)009<0209:ASHIPS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., and J. Kaplan, 1999: An updated Statistical Hurricane Intensity Prediction Scheme (SHIPS) for the Atlantic and eastern North Pacific basins. Wea. Forecasting, 14, 326337, doi:10.1175/1520-0434(1999)014<0326:AUSHIP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • DeMaria, M., M. Mainelli, L. K. Shay, J. A. Knaff, and J. Kaplan, 2005: Further improvements to the Statistical Hurricane Intensity Prediction Scheme (SHIPS). Wea. Forecasting, 20, 531543, doi:10.1175/WAF862.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1986: An air–sea interaction theory for tropical cyclones. Part I: Steady-state maintenance. J. Atmos. Sci., 43, 585605, doi:10.1175/1520-0469(1986)043<0585:AASITF>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hendricks, E. A., M. S. Peng, B. Fu, and T. Li, 2010: Quantifying environmental control on tropical cyclone intensity change. Mon. Wea. Rev., 138, 32433271, doi:10.1175/2010MWR3185.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holliday, C. R., and A. H. Thompson, 1979: Climatological characteristics of rapidly intensifying typhoons. Mon. Wea. Rev., 107, 10221034, doi:10.1175/1520-0493(1979)107<1022:CCORIT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • JAXA, 2016a: JAXA global rainfall watch. Accessed 18 November 2016. [Available online at http://sharaku.eorc.jaxa.jp/GSMaP/index.htm.]

  • JAXA, 2016b: JAXA realtime rainfall watch. Accessed 18 November 2016. [Available online at http://sharaku.eorc.jaxa.jp/GSMaP_NOW/index.htm.]

  • Jiang, H., 2012: The relationship between tropical cyclone intensity change and the strength of inner-core convection. Mon. Wea. Rev., 140, 11641176, doi:10.1175/MWR-D-11-00134.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jiang, H., and E. M. Ramirez, 2013: Necessary conditions for tropical cyclone rapid intensification as derived from 11 years of TRMM data. J. Climate, 26, 64596470, doi:10.1175/JCLI-D-12-00432.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Jones, S. C., 1995: The evolution of vortices in vertical shear. I: Initially barotropic vortices. Quart. J. Roy. Meteor. Soc., 121, 821851, doi:10.1002/qj.49712152406.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Joyce, R. J., J. E. Janowiak, P. A. Arkin, and P. Xie, 2004: CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high spatial and temporal resolution. J. Hydrometeor., 5, 487503, doi:10.1175/1525-7541(2004)005<0487:CAMTPG>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kaplan, J., and M. DeMaria, 2003: Large-scale characteristics of rapidly intensifying tropical cyclones in the North Atlantic basin. Wea. Forecasting, 18, 10931108, doi:10.1175/1520-0434(2003)018<1093:LCORIT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kaplan, J., M. DeMaria, and J. A. Knaff, 2010: A revised tropical cyclone rapid intensification index for the Atlantic and eastern North Pacific basins. Wea. Forecasting, 25, 220241, doi:10.1175/2009WAF2222280.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kieper, M., and H. Jiang, 2012: Predicting tropical cyclone rapid intensification using the 37 GHz ring pattern identified from passive microwave measurements. Geophys. Res. Lett., 39, L13804, doi:10.1029/2012GL052115.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kieu, C., V. Tallapragada, and W. Hogsett, 2014: Vertical structure of tropical cyclones at onset of the rapid intensification in the HWRF model. Geophys. Res. Lett., 41, 32983306, doi:10.1002/2014GL059584.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kubota, T., and Coauthors, 2007: Global precipitation map using satellite-borne microwave radiometers by the GSMaP project: Production and validation. IEEE Trans. Geosci. Remote Sens., 45, 22592275, doi:10.1109/TGRS.2007.895337.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kubota, T., S. Shige, K. Aonashi, and K. Okamoto, 2009: Development of nonuniform beamfilling correction method in rainfall retrievals for passive microwave radiometers over ocean using TRMM observations. J. Meteor. Soc. Japan, 87A, 153164, doi:10.2151/jmsj.87A.153.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miyamoto, Y., and T. Takemi, 2013: A transition mechanism for the axisymmetric spontaneous intensification of tropical cyclones. J. Atmos. Sci., 70, 112129, doi:10.1175/JAS-D-11-0285.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miyamoto, Y., and T. Takemi, 2015: A triggering mechanism for rapid intensification of tropical cyclones. J. Atmos. Sci., 72, 26662681, doi:10.1175/JAS-D-14-0193.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molinari, J., and D. Vollaro, 2010: Rapid intensification of a sheared tropical storm. Mon. Wea. Rev., 138, 38693885, doi:10.1175/2010MWR3378.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Molinari, J., D. Vollaro, and K. L. Corbosiero, 2004: Tropical cyclone formation in a sheared environment: A case study. J. Atmos. Sci., 61, 24932509, doi:10.1175/JAS3291.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nolan, D. S., and L. D. Grasso, 2003: Nonhydrostatic, three-dimensional perturbations to balanced, hurricane-like vortices. Part II: Symmetric response and nonlinear simulations. J. Atmos. Sci., 60, 27172745, doi:10.1175/1520-0469(2003)060<2717:NTPTBH>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nolan, D. S., Y. Moon, and D. P. Stern, 2007: Tropical cyclone intensification from asymmetric convection: Energetics and efficiency. J. Atmos. Sci., 64, 33773405, doi:10.1175/JAS3988.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pendergrass, A. G., and H. E. Willoughby, 2009: Diabatically induced secondary flows in tropical cyclones. Part I: Quasi-steady forcing. Mon. Wea. Rev., 137, 805821, doi:10.1175/2008MWR2657.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reasor, P. D., and M. T. Montgomery, 2001: Three-dimensional alignment and corotation of weak, TC-like vortices via linear vortex Rossby waves. J. Atmos. Sci., 58, 23062330, doi:10.1175/1520-0469(2001)058<2306:TDAACO>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reasor, P. D., M. T. Montgomery, and L. D. Grasso, 2004: A new look at the problem of tropical cyclones in vertical shear flow: Vortex resiliency. J. Atmos. Sci., 61, 322, doi:10.1175/1520-0469(2004)061<0003:ANLATP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Reasor, P. D., R. Rogers, and S. Lorsolo, 2013: Environmental flow impacts on tropical cyclone structure diagnosed from airborne Doppler radar composites. Mon. Wea. Rev., 141, 29492969, doi:10.1175/MWR-D-12-00334.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Riemer, M., M. T. Montgomery, and M. E. Nicholls, 2010: A new paradigm for intensity modification of tropical cyclones: Thermodynamic impact of vertical wind shear on the inflow layer. Atmos. Chem. Phys., 10, 31633188, doi:10.5194/acp-10-3163-2010.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rogers, R. F., P. Reasor, and S. Lorsolo, 2013: Airborne Doppler observations of the inner-core structural differences between intensifying and steady-state tropical cyclones. Mon. Wea. Rev., 141, 29702991, doi:10.1175/MWR-D-12-00357.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rogers, R. F., P. Reasor, and J. Zhang, 2015: Multiscale structure and evolution of Hurricane Earl (2010) during rapid intensification. Mon. Wea. Rev., 143, 536562, doi:10.1175/MWR-D-14-00175.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schubert, W. H., and J. J. Hack, 1982: Inertial stability and tropical cyclone development. J. Atmos. Sci., 39, 16871697, doi:10.1175/1520-0469(1982)039<1687:ISATCD>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shapiro, L., and H. E. Willoughby, 1982: The response of balanced hurricanes to local sources of heat and momentum. J. Atmos. Sci., 39, 378394, doi:10.1175/1520-0469(1982)039<0378:TROBHT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shige, S., and Coauthors, 2009: The GSMaP precipitation retrieval algorithm for microwave sounders. Part I: Over-ocean algorithm. IEEE Trans. Geosci. Remote Sens., 47, 30843097, doi:10.1109/TGRS.2009.2019954.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stern, D. P., J. L. Vigh, D. S. Nolan, and F. Zhang, 2015: Revisiting the relationship between eyewall contraction and intensification. J. Atmos. Sci., 72, 12831306, doi:10.1175/JAS-D-14-0261.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ushio, T., and Coauthors, 2009: A Kalman filter approach to the Global Satellite Mapping of Precipitation (GSMaP) from combined passive microwave and infrared radiometric data. J. Meteor. Soc. Japan, 87A, 137151, doi:10.2151/jmsj.87A.137.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vigh, J. L., and W. H. Schubert, 2009: Rapid development of the tropical cyclone warm core. J. Atmos. Sci., 66, 33353350, doi:10.1175/2009JAS3092.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zagrodnik, J. P., and H. Jiang, 2014: Rainfall, convection, and latent heating distributions in rapidly intensifying tropical cyclones. J. Atmos. Sci., 71, 27892809, doi:10.1175/JAS-D-13-0314.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
Save